testing alternatives for salt wedge management in … · the greek territory through the straits of...
TRANSCRIPT
INTRODUCTIONEstuaries form the transition zone between the
inland freshwater and seawater. Due to their posi-
tion they are characterized by some unique fea-
tures and properties, one of which is the intrusion
of seawater upstream of the river mouth (when-
ever river topography is below mean sea level),
mainly caused by its higher density. Thus, one of
the problems of water management in estuarine
areas is the control of salinity, as salinity limits
have been assessed for the use of water for vari-
ous purposes, including irrigation.
Several methods have been proposed to control
or prevent salt intrusion. These include, among
Global Nest: the Int. J. Vol 5, No 2, pp 107-118, 2003
Copyright© 2003 GLOBAL NEST
Printed in Greece. All rights reserved
TESTING ALTERNATIVES FOR SALT WEDGE MANAGEMENT
IN AN ESTUARY WITH THE USE OF MONITORING
AND A MATHEMATICAL MODEL
1 Democritus University of Thrace, School of Engineering
Department of Environmental Engineering
Laboratory of Ecological Engineering and Technology
67100 Xanthi, Greece2 National Agricultural Research Foundation, Fisheries Research Institute
64007 Nea Peramos, Kavala, Greece
*to whom all correspondence should be addressed:
fax: +30-25410 78113
e-mail: [email protected]
ABSTRACTThe intrusion of salt wedge in rivers is a natural phenomenon, which occurs in many estuaries. Saline
water tends to propagate upstream from the river mouth, due to the limited freshwater and the tidal
and density currents developed, resulting in deterioration of water quality in the lower river reach.
Several methods to control the salt wedge have been employed, including the construction of inflatable
dams or gates. A promising method of control is the use of an air curtain. In this study, a two-dimen-
sional, laterally averaged numerical model has been developed to describe salt wedge intrusion. This
model provided necessary hydraulic parameters, which were used in air curtain design theory to evalu-
ate the application of the air curtain method in a particular estuary system. The application takes place
in the estuary of Strymon River in Northern Greece, where the limited discharge of freshwater, main-
ly caused by the construction of Kerkini dam, results in the creation and upstream intrusion of a salt
wedge in the summertime, affecting water quality and making water unsuitable for irrigation uses.
KEYWORDS: saline intrusion, salt wedge, estuary management, air curtain, numerical modeling,
Strymon Estuary.
K.I. HARALAMBIDOU1
G.K. SYLAIOS1,2
V.A. TSIHRINTZIS1,*
Selected from papers presented at the 8th Conference
on Environmental Science and Technology,
8 - 10 September 2003, Lemnos, Greece.
14-Haralambidou.qxd 9/11/2005 10:04 Page 107
others the construction of small movable dams at
the river mouth and the increase of the roughness
of the river bottom. One proposed method is the
use of an air curtain, which may reduce or prevent
totally the intrusion of the wedge. According to
this method, compressed air is pumped perpen-
dicularly to the flow from a perforated pipe,
placed on the river bed across the channel, form-
ing a vertical air curtain which acts as a wall to the
intrusion. The use of the air curtain has been
studied through limited laboratory experiments
and has been applied in several navigation locks
(e.g., Abraham and Burgh, 1964; Raman and
Arbuckle, 1989, Hamilton et al., 2001).
This paper presents an application of the air cur-
tain technology in the estuary of Strymon River,
Northern Greece, where the limited discharge of
freshwater, mainly caused by the upstream con-
struction of Kerkini dam, results in a salt wedge
intrusion, affecting water quality and making
water unsuitable for irrigation uses. To deal with
the problem of the presence of salt wedge in
Strymon, an extended study has been undertaken,
which includes three major parts: (1) design and
application of an extensive field data collection
program along the river estuary (Haralambidou
et al., 2003a); (2) development of a numerical
model, which is used to test alternatives to control
the salt wedge intrusion (Haralambidou et al.,
2003b); and (3) testing of the effectiveness of the
air curtain method with the use of the model
(Haralambidou et al., 2003c). In this paper, the
numerical model and its use in testing the applic-
ability of the air curtain method in Strymon River
Estuary are presented.
MATERIALS AND METHODSNumerical model descriptionA two-dimensional, laterally-intregrated, explicit,
finite-difference numerical model has been devel-
oped to describe the characteristics of the salt
wedge intrusion in the estuary of Strymon River.
The governing equations of the model are based
on the principles of conservation of volume,
momentum and mass.
Considering hydrostatic pressure, the Boussinesq
approximation, and a Cartesian coordinate sys-
tem with the origin at the sea surface of the estu-
ary mouth (mean sea level), in which the x-axis is
directed upstream and the z-axis is directed
upward, the flow governing equations read:
The laterally integrated continuity equation:
(1)
The vertically-integrated continuity equation:
(2)
The momentum conservation equation:
(3)
where:
(4)
The salt balance equation:
(5)
The equation of state:
(6)
where: x and z = horizontal and vertical spatial
coordinates, cm; t = time, s; ρ(x,z,t) = water den-
sity, g cm-3; u(x, z, t) and w(x, z, t) = horizontal
and vertical laterally-averaged velocities, cm/s;
B(x, z) = channel width, cm; Bç = width at the
free surface, cm; g = acceleration due to gravity,
cm s-2; η(x, t) = sea surface displacement, cm; d =
water depth below mean sea level, cm; NX and NZ
= horizontal and vertical eddy viscosity coeffi-
cients, cm2 s-1; KX and KZ = horizontal and verti-
cal eddy diffusivity coefficients, cm2 s-1; k =
empirically determined drag coefficient, which is
expressed as: , where n the Man-
ning roughness coefficient, s cm-1/3; S(x, z, t) =
108 HARALAMBIDOU et al.
14-Haralambidou.qxd 9/11/2005 10:04 Page 108
salinity, psu; ρo = 0.9995 g cm-3 (density of fresh-
water); and á, â = constants, functions of tem-
perature. For water temperature of 15oC, á =
1.00059 and â = 7.57x10-4.
A first order turbulence closure scheme was used
for the parameterization of KZ and NZ eddy coef-
ficients, considering their dependence on the
local Richardson number Ri. In the present
model, the formulations of KZ and NZ are
(Blumberg, 1975, 1977):
(7)
for (8)
for (9)
where: k1 = 0.5 cm2 s-1, ãc = Kc/Nc for the critical
condition of stratification (Ric = 10, ãc ∼0.05-1.2)
(Yamada, 1975).
Instantaneous values of KZ and NZ are strongly
dependent on flow velocity, Manning's roughness
coefficient and vertical stratification (Sylaios,
1994; Parissis et al., 2001).
To eliminate non-linear instabilities in the model
results, an artificial horizontal viscosity term was
introduced (Smagorinsky, 1963). Horizontal eddy
viscosity and diffusivity coefficients KX and NX
were computed by the model, based upon the for-
mulation of:
where: c an adjustable constant and Äx the hori-
zontal grid spacing. Therefore, from this artificial
viscosity term, wherever the horizontal gradients
of velocity are large, the eddy viscosity will
become large.
The air curtain methodSeveral methods have been proposed to control
or prevent intrusion. One of them is the imple-
mentation of an air curtain (Abraham and Burgh,
1964; Tuin et al., 1991; Kerstma et al., 1994;
Hamilton et al., 2001). Nakai and Arita (2002)
presented experimental results on control of salt
wedge intrusion in rivers using this method. Their
theory is used in this study, based on the following
short description.
The salt wedge is influenced by an air curtain in
various ways. Nakai and Arita (2002) classified
the salt wedge behavior in three types, as illus-
trated in Figure 1. Under Type I, salt water
intrudes into the upstream side over the air cur-
tain along the channel bottom. This type appears
when the kinetic energy provided by the buoyan-
cy due to the air curtain is small. Under Type II,
saltwater is controlled downstream of the curtain
by the strong upward flow. Under Type III, salt-
water is also controlled downstream of the air
curtain, but some portion enters the upstream
side along the channel bottom. Obviously the pre-
ferred condition is Type II.
Nakai and Arita (2002) found that external forces
dominating the operation of the system are the
following three: the buoyancy due to the air cur-
tain, A, the intrusion force of the salt wedge, B,
and the inertial force of the freshwater flow, R.
All parameters can be written in velocity dimen-
sions (i.e., m/s), as follows (Nakai and Arita,
2002):
A = (qa g)1/3 ;
B = (g′ ha)1/2 ; (10)
R = g f /d
where: qa = discharge per unit width of air; g = grav-
itational acceleration; g ′ = reduced gravitational
acceleration [ ]; ñs and ñf are the
densities of saltwater and freshwater, respective-
ly; ha is the salt wedge thickness at the position of
the air curtain maker, in the absence of the air
curtain; qf = discharge per unit width of freshwa-
ter flow; and d = total water depth. The behavior
of a salt wedge around an air curtain depends on
both ratios A/B and A/R, as shown in Figure 2.
Description of the study areaSalt wedge intrusion is observed at the estuary of
Strymon River in Northern Greece. Strymon is a
transboundary river, with a total channel length of
315 km and a total drainage area of 18,329 km2,
39.8% of which belongs to Greek territory
(Hatzigiannakis, 2000). With a NW-SE direction,
109TESTING ALTERNATIVES FOR SALT WEDGE MANAGEMENT IN AN ESTUARY
14-Haralambidou.qxd 9/11/2005 10:04 Page 109
Strymon River enters the Serres drainage area in
the Greek territory through the straits of Roupel,
passes the Kerkini dam located about 70 km
upstream of the estuary and flows into the
Strymonikos gulf in the North Aegean Sea
(Figure 3).
The present study involves the estuary system of
the river and particularly the area upstream of the
river mouth up to the straits of Amphipolis, a dis-
tance of approximately 6-8 km. The Strymon
Estuary is of the coastal plain type, forced at its
mouth by a micro-tidal M2-sinusoidal wave, with a
spring range of approximately 0.45 m and a neap
range of 0.10 m. The mean depth of river water in
the study area is approximately 3 m, varying in the
range 2-5 m, and the mean width is 64 m, varying
in the range 40-90 m. An underwater sill also
exists at the mouth of the estuary, approximately
200 m long and 0.2-0.8 m deep (Vouvalides,
1998). The river discharge varies in the range 0-
120 m3 s-1 (Hatzigiannakis et al., 2000) due to the
presence of the Kerkini dam in the pathway of the
river.
The wedge develops mostly in the summer, when
there is a great demand for irrigation water for
the nearby agricultural fields, as the freshwater
discharge from the dam is minimal. A critical
point in the estuary is at a distance of approxi-
mately 3 km from the mouth, where a pumping
station, which pumps water for irrigation, has
been installed. The pump is equipped with a salin-
ity sensor, which interrupts pumping when the
water quality is unsuitable for irrigation, i.e.,
when the salinity is greater than 15 psu.
Therefore, in order for the pump to operate, the
salt wedge should not extend beyond 3 km from
the mouth.
The field survey was conducted in winter of 2002
and in spring and summer of 2003. Three prelim-
inary sampling campaigns were conducted in win-
ter and spring for the study of seasonal variations
(Haralambidou et al., 2003a; 2004). The saline
wedge was absent at this period. The summer
campaigns were more intense because of the
presence of the saline wedge. Five sampling cam-
paigns were made in summer, namely one in
June, two in July and two in August. In most cases
in the summer the sampling lasted about 12h.
Temperature, conductivity, salinity, pH, dissolved
oxygen and velocity were measured along the cen-
terline of the estuary (Figure 3). Data were col-
lected every 0.25 m from surface to bottom and
110 HARALAMBIDOU et al.
Figure 2. The dependency of the behavior of the salt wedge on ratios A/B and A/R (Nakai and Arita, 2002).
Figure 1. Flow type classification according to Nakai and Arita (2002).
14-Haralambidou.qxd 9/11/2005 10:04 Page 110
profiles of the above mentioned parameters were
obtained using conductivity, pH and dissolved
oxygen meters (WTW). The total number of sta-
tions along the estuary was variable, as the length
of the saline wedge was different based on river
discharge and tide.
NUMERICAL EXPERIMENTS AND RESULTSModel resultsEquations (1) to (6) were solved using the
method of finite-differences. The computer code
was written in FORTRAN 77 and was compiled
using PROSPERO FORTRAN. The grid was in
the vertical plane, with spacings set at Äx = 1000
m horizontally and Äz = 0.5 m vertically, and a
time step increment of Ät = 15 sec. A run in time
of two tidal periods was found adequate for the
tidal regime and the vertical profiles of velocity
and salinity to become established. The results of
the second period were used in the following
analysis. It was assumed that there was no salt flux
through the surface or bottom and that there was
no wind stress on the surface. The initial distribu-
tion of the salinity field was defined by the moni-
toring surveys with value at the river mouth of
approximately 35 psu and value at the head water
of 0 psu.
The parameters of L15 and L30 were introduced to
define the salinity intrusion length along the up-
estuary direction. The parameters L15 and L30
represent the salt intrusion length defined by the
bottom position of the 15 psu and 30 psu-isoha-
lines, respectiverly, measured from the river
mouth (Tuin et al., 1991). The model was used to
investigate the influence of the river discharge,
the tidal amplitude, the bottom Manning's rough-
ness coefficient and the topography of the estuary
on the salinity structure in the Strymon Estuary.
The conditions simulated by the model are sum-
marized in Table 1. Figure 5 illustrates a repre-
sentative salinity profile under the typical summer
flow condition of Q = 15 m3 s-1, with a spring tidal
111TESTING ALTERNATIVES FOR SALT WEDGE MANAGEMENT IN AN ESTUARY
Figure 3. Map of Strymon River system and its location in Greek territory.
14-Haralambidou.qxd 9/11/2005 10:04 Page 111
range of DE = 45 cm and Manning's roughness
coefficient n = 0.050.
Effect of river discharge: To study the responses of
salinity distribution to varying river discharge, the
model was run for river discharge values Q=1 m3
s-1 to Q=30 m3 s-1, with tested values of 1, 5, 15, 20
and 30 m3 s-1, and with a spring tidal range of
DE=45 cm and a Manning's roughness coeffi-
cient n = 0.050 (Table 1). Figure 6a shows the
salinity intrusion lengths L15 and L30 as function
of river discharge, Q. It is apparent that the mag-
nitude of the salinity intrusion length depends
inversely on the magnitude of the river discharge,
as expected. It is also seen that the freshwater dis-
charge is a quite sensitive parameter in control-
ling salt wedge intrusion.
Effect of tidal amplitude: The model was also run
for different tide regimes. The neap tide, the
mean and the spring tides have been tested, with
amplitudes DE of 10 cm, 18 cm and 45 cm. The
river discharge was set to Q = 15 m3 s-1, and the
Manning's roughness coefficient to n = 0.050
(Table 1). The results are presented in Figure 6b.
It is apparent that there is a nearly exponential
increase of the salinity intrusion length with the
increase of tidal amplitude. For the given fresh-
water discharge, it is also seen that the neap tide
under these conditions the 30 psu isohaline does
not enter the river (L30 = 0) but the 15 psu isoha-
line progresses more than 4 km up-estuary.
Again, the tidal amplitude is a quite sensitive
parameter in controlling salt wedge intrusion.
Effect of Manning's roughness coefficient: The
effect of varying the Manning's roughness coeffi-
cient n was tried by making runs using n values
from 0.020 to 0.065, with tested values of 0.020,
0.035, 0.040, 0.050, 0.060 and 0.065, with constant
values of river discharge (Q = 15 m3 s-1) and tidal
amplitude (DE = 45 cm) (Table 1). The results
are shown in Figure 6c. It is apparent that a small
decrease of saline intrusion takes place as the
roughness coefficient increases.
Effect of the river mouth topography: The model
was run for different widths and underwater sill
heights at the mouth. The widths W have been
tested in the range from W=50 m to W=80 m,
with tested values of 50, 60, 70 and 80 m. In
another set of runs, the underwater sill heights at
the river mouth SH varied from SH=0.5 m to
SH=2 m, with tested values of 0.5, 1, 1.5 and 2 m
(i.e., the sill was raised). The parameters of river
discharge (Q=15 m3 s-1) and tidal amplitude
(DE=45cm) remained constant throughout these
tests (Table 1). A small increase in the length of
salinity intrusion was observed as the width
increased, which was more obvious on the 15 psu-
112 HARALAMBIDOU et al.
Table 1. Summary of conditions of numerical model experiments.
Case 1: Case 2:Case 3:
Case 4: Case 5:
Typical Effect of Effect ofEffect of
Effect of Effect ofParameter
Conditions Freshwater TidalManning's
Mouth Sill
Discharge AmplitudeRoughness
Width HeightCoefficient
FreshwaterVarying:
Discharge 151 - 30
15 15 15 15
(m3 s-1)
TidalVarying:
Amplitude 18 4510 - 45
45 45 45
(m)
Manning'sVarying:
Roughness 0.050 0.050 0.0500.020 - 0.065
0.050 0.050
Coefficient
Mouth80 80 80 80
Varying:80
Width (m) 50 - 80
Sill Height2.5 2.5 2.5 2.5 2.5
Varying:
(m) 0.5 - 2.0
14-Haralambidou.qxd 9/11/2005 10:04 Page 112
isohaline than the 30 psu-isohaline. On the con-
trary, no significant change was observed as the
underwater sill height at the mouth rose (Figure
6d, 6e).
Air curtain resultsThe model was used to obtain the values of den-
sity of saltwater and freshwater for different val-
ues of river discharge Q, tidal amplitude DE,
Manning's roughness coefficient n, widths W and
sill height SH at the mouth of the estuary. The
conditions simulated by the model and tested
using the air curtain, are summarized in Table 2.
Flow types (i.e., I, II or III according to Figures 3
and 4) for the conditions mentioned above result-
ed from the calculation of the ratios A/R and A/B
(Eqs. 7). Figure 7 presents some of these results
based on the number of observations made. The
following are observed:
• As the river freshwater discharge increases the
flow type moves from Type I to II.
• For the same values of river discharge, an
increase in the air flow qa has as a conse-
quence the elimination of Type I condition
and the appearance of Type II. Type III
appears at minimum tested river discharge of
Q = 15 m3 s-1.
• There is no significant effect on flow type for
changes in the Manning's roughness coeffi-
cient, sill height at the river mouth and width
of the channel. However, there is an impact on
the flow type when the discharge of the air
increases. In particular for qa = 0.01 m3 s-1 per
unit width all conditions of Manning's rough-
ness coefficient tested are of Type I and for qa
= 0.05 m3 s-1 per unit width most conditions
are of Type III. Exactly the same happens in
the cases of varying sill height and width of the
channel.
• In relation to tidal amplitude, the neap tide
condition appears to be out of the range of the
diagram, whereas the spring tide condition is
of Type I and the mean tide condition is of
Type II. As qa rises mean tide condition
remains of Type II and spring tide condition
moves from Type I to Type III.
PROPOSALS FOR THE MANAGEMENT OF THESALT WEDGEFrom the sensitivity analysis of the model it is
obvious that:
1. River freshwater discharge is the most sensi-
tive parameter. Thus, if no other measures are
taken to solve the problem, a minimum con-
stant flow of 30 m3 s-1 should be negotiated by
authorities and be released from Kerkini dam
at all times. This flow would keep the 15 psu
isohaline downstream of the pumping station
even during spring tides.
2. The tidal amplitude is also a very sensitive
parameter, affecting salt wedge intrusion.
3. Manning's roughness coefficient has only a
minor effect on the propagation of the salt
wedge. Thus, roughening the riverbed by
installing artificial concrete blocks or dumping
rock and crushed stone would be a rather
expensive and ineffective solution. This could,
however, be done in conjunction with another
measure.
4. Reducing the width of the river mouth has
some minor impact on the salt wedge propa-
gation. It could be done relatively inexpensive-
ly, but since it cannot eliminate totally the
problem, it should be done in conjunction with
another measure.
5. Raising also the sill height at the mouth did
not show to have a significant impact on the
propagation of the salt wedge. Furthermore,
this measure would create problems with
small fishing boats passing through the mouth.
Thus, this should not be an option, unless, as
mentioned above, the tide is totally blocked,
for example, with an inflatable dam.
From the application of the air curtain method it
is apparent that:
1. When the river discharge is greater that Q = 30
m3 s-1, a discharge of air qa of the order of 0.01
m3 s-1 per unit width is sufficient to produce
flow of Type II. On the contrary, for river dis-
charge less that Q = 30 m3 s-1 an increase in qa
improves the situation from Type I to Type II.
2. In relation to varying Manning's roughness
coefficient and characteristics of the topogra-
phy of the estuary, an increase in qa does not
alter the type of the flow and is considered to
be meaningless.
3. Under spring tide conditions high values of qa
are more effective, as they alter the condition
from Type I to Type III. However, under
mean tide condition there is no need to alter
the qa value.
113TESTING ALTERNATIVES FOR SALT WEDGE MANAGEMENT IN AN ESTUARY
14-Haralambidou.qxd 9/11/2005 10:04 Page 113
CONCLUSIONSA two-dimensional, laterally-integrated, explicit
finite-difference numerical model was developed
to describe salt wedge dynamics in Strymon River
estuary. This model was applied successfully and
proved to be an effective tool in testing alterna-
tive management scenarios for controlling salt
wedge intrusion in this estuary. The effect of river
freshwater discharge, tidal amplitude, bottom
Manning's roughness coefficient, river mouth
width and underwater sill height at the river
mouth was also studied. Finally, the model pro-
vided the required minimum air flow value for air
curtain technology application to stop salt wedge
upstream progression under given river flow and
tide amplitude conditions.
114 HARALAMBIDOU et al.
Table 2. Summary of conditions tested using the air curtain method for density of fresh water ñf = 1.005396 g
cm-3, total water depth d = 3 m and testing discharges of air qa = 0.01 and 0.05 m3 s-1 per unit width.
Parameter qf(m3/s per unit width) ñ
s(g/cm3)
River Discharge Q (m3/s)
15 0.229 1.0224
20 0.305 1.0179
30 0.457 1.0098
40 0.610 1.0084
50 0.762 1.0080
Tidal Amplitude DE (cm)
10 0.229 1.0055
18 0.229 1.0101
45 0.229 1.0224
Manning's Roughness Coefficient n
0.020 0.229 1.0290
0.030 0.229 1.0282
0.040 0.229 1.0258
0.050 0.229 1.0224
0.060 0.229 1.0190
0.065 0.229 1.0174
Width of Estuary Mouth W (m)
50 0.229 1.0199
60 0.229 1.0218
70 0.229 1.0225
80 0.229 1.0224
Underwater Sill Height SH (m)
0.5 0.229 1.0241
1.0 0.229 1.0236
1.5 0.229 1.0229
2.0 0.229 1.0224
qf = discharge of freshwater flow per unit width; ñs = density of saltwater
14-Haralambidou.qxd 9/11/2005 10:04 Page 114
115TESTING ALTERNATIVES FOR SALT WEDGE MANAGEMENT IN AN ESTUARY
Figure 5. Longitudinal salinity distribution computed by the model under river discharge of Q = 15 m3/s at
spring tide range DE = 45 cm and with Manning's roughness coefficient n = 0.050. The lines are iso-
halines and the labels represent salinities in psu. The x-axis represents number of Äx's (Äx = 1000 m)
and the z-axis number of Äz's (Äz = 0.5m).
Figure 4. Longitudinal profiles of salinity upstream of the mouth of Strymon River measured on 31-08-2003 at:
(a) 11:00 am; (b) 13:00 pm; (c) 15:30 pm; and (d) 18:00 pm. The x-axis is in m and the z-axis is in cm.
Isohaline labels are in psu.
14-Haralambidou.qxd 9/11/2005 10:04 Page 115
116 HARALAMBIDOU et al.
Figure 6. The relation of salinity intrusion length, LS to: (a) the river discharge Q; (b) the tidal amplitude DE;
(c) the Manning's roughness coefficient n; (d) the width of the estuary mouth W; and (e) the under-
water sill height at the estuary mouth SH.
14-Haralambidou.qxd 9/11/2005 10:04 Page 116
117TESTING ALTERNATIVES FOR SALT WEDGE MANAGEMENT IN AN ESTUARY
Figure 7. Diagrams of various conditions for two flowrates of air.
14-Haralambidou.qxd 9/11/2005 10:04 Page 117
118 HARALAMBIDOU et al.
REFERENCESAbraham, G. and Burgh, P.v.d. (1964), Pneumatic reduction of salt intrusion through locks, Journal of Hydraulics
Division, ASCE, 90(HY1), 83-119.
Blumberg, A.F. (1975), A numerical investigation into the dynamics of estuarine circulation, Technical Report 91,
Chesapeake Bay Institute, The Johns Hopkins University, Baltimore.
Blumberg, A.F. (1977), Numerical model of estuarine circulation, Journal of Hydraulics Division, ASCE,
103(HY3), 295-310.
Hamilton, D.P., Chan, T., Robb, M.S., Pattiaratch, C.B. and Herzfeld M. (2001), The hydrology of the upper Swan
River Estuary with focus on an artificial destratification trial, Hydrological Processes, 15, 2465-2480.
Haralambidou, K., Sylaios, G.K., Tsihrintzis, V.A. and Akratos, C. (2003a), A monitoring program for the study
and management of the Strymon River Estuary, Proc. 9th National Conference of the Hellenic
Hydrotechnical Society, April 2-5, Thessaloniki, Greece, pp. 513-520 (in Greek).
Haralambidou, K., Sylaios, G.K. and Tsihrintzis, V.A. (2003b), Development of a numerical model to test alter-
natives to control saline wedge intrusion in an estuary, Proc. of the XXX IAHR Congress, Theme A -
Coastal Environment: Processes and Integrated Management, August 24-29, Thessaloniki, Greece, pp.
107-114.
Haralambidou, K.I., Tsihrintzis, V.A. and Sylaios, G.K. (2003c), Control of saline wedge intrusion in the estuary
of Strymonas River using an air curtain, Proc. of the 8th Conference on Environmental Science and
Technology, September 8-10, Porto Myrina Palace, Lemnos, Greece, Vol. A, pp. 302-309.
Haralambidou, K., Tsihrintzis, V.A., Sylaios, G.K. and Akratos, C. (2004). Seasonal and spatial characteristics of
water quality in the estuary of Strymon River, Proc. of the International Conference on Protection and
Restoration of the Environment VII, June 28-July 1, 2004, Mykonos, Greece, CD-Rom Section III
(Surface Water Flow and Quality), #3.
Hatzigiannakis, S. (2000), Hydrology of Strymon River, Report on the description of the coastal zone of
Strymonikos and Ierissos Gulfs, (eds. E. Koutrakis and E. Lazaridou), Fisheries Research Institute -
NAGREF, Greek Biotope/Wetland Center, Nea Peramos, Kavala, Greece, pp. 5-32 (in Greek).
Kerstma, J., Kolkman, P.A., Regeling, H.J. and Venis, W.A. (1994), Water quality contol at ship locks, prevention
on salt and fresh water exchange, Balkema, Netherlands.
Nakai, M. and Arita, M. (2002), An experimental study on the prevention of saline wedge intrusion by an air cur-
tain in rivers, Journal of Hydraulic Research, 40(3), 333-339.
Parissis, A., Sylaios, G. and Tsihrintzis, V.A. (2001), A numerical model for the study of salt intrusion at
Strymonas river mouth, Northern Greece, Proc. of 1st International Conference on Ecological Protection
of the Planet Earth, (eds. V.A. Tsihrintzis and P. Tsalides), June 5-8, Xanthi, Greece, pp. 281-288.
Raman, K. and Arbuckle, B. (1989), Long-term destratification in an Illinois Lake, Journal of American Water
Works Association, 81, 66-71.
Smagorinsky, J. (1963), General circulation experiments with the primitive equations, Monthly Weather Review, 91,
99-164.
Sylaios, G. (1994), A numerical investigation into the dynamics of a partially-mixed estuary, Ph.D. Thesis, Dept. of
Oceanography, Faculty of Science, University of Southampton, U.K.
Sylaios, G. (1999), Physical oceanography of Strymonikos Gulf and Gulf of Ierissos, Final Report on the description
of the coastal zone of Strymonikos and Ierissos Gulfs, (eds. E. Koutrakis and E. Lazaridou), Fisheries
Research Institute - NAGREF, Greek Biotope/Wetland Center, Nea Peramos, Kavala, Greece, pp.
119-218 (in Greek).
Tuin, H.v.d.T., Le, H.-T., Mikhailov, V., Roelfzema, A. and Volker, A. (1991), Guidelines on the study of seawater
intrusion into rivers, UNESCO, France.
Vouvalides, K. (1998), Morphological, sedimentological and oceanographic processes and human impacts on the evo-
lution of the Strymon River, Ph.D. Thesis, Dept. of Geology, Aristotle University of Thessaloniki,
Greece (in Greek).
Yamada, T. (1975), The critical Richardson number and the ratio of the eddy transport coefficients obtained from
a turbulence model, Journal of Atmospheric Science, 32, 926-933.
14-Haralambidou.qxd 9/11/2005 10:04 Page 118